Why document loading is the RAG foundation in LangChain - Test Your Understanding

Practice - 5 Tasks

1fill in blank

easy

Complete the code to load documents using LangChain's loader.

LangChain

from langchain.document_loaders import TextLoader
docs = TextLoader('[1]').load()

Drag options to blanks, or click blank then click option'

Amodel.pkl

Bdata.txt

Cconfig.json

Doutput.csv

Attempts:

3 left

💡 Hint

Common Mistakes

Using a model or config file instead of a text document.

The TextLoader loads text documents from a file. Here, 'data.txt' is the correct file to load documents.

2fill in blank

medium

Complete the code to split documents into chunks for better retrieval.

LangChain

from langchain.text_splitter import RecursiveCharacterTextSplitter
splitter = RecursiveCharacterTextSplitter(chunk_size=[1], chunk_overlap=20)
docs = splitter.split_documents(docs)

Drag options to blanks, or click blank then click option'

A1000

B10

C5000

D50

Attempts:

3 left

💡 Hint

Common Mistakes

Using too small or too large chunk sizes that hurt retrieval quality.

Chunk size of 1000 characters is a common choice to balance chunk size and overlap for retrieval.

3fill in blank

hard

Fix the error in the code to create a vector store from documents.

LangChain

from langchain.vectorstores import FAISS
from langchain.embeddings.openai import OpenAIEmbeddings
embeddings = OpenAIEmbeddings()
vectorstore = FAISS.from_documents(docs, [1])

Drag options to blanks, or click blank then click option'

AOpenAIEmbeddings

BFAISS

Cembeddings

Ddocs

Attempts:

3 left

💡 Hint

Common Mistakes

Passing documents or class names instead of embeddings instance.

The FAISS vector store requires the embeddings object to convert documents into vectors.

4fill in blank

hard

Fill both blanks to create a RetrievalQA chain with the vector store retriever.

LangChain

from langchain.chains import RetrievalQA
retriever = vectorstore.[1]()
qa_chain = RetrievalQA.from_chain_type(llm=llm, retriever=[2])

Drag options to blanks, or click blank then click option'

Aas_retriever

Bget_retriever

Cretriever

Dretrieve

Attempts:

3 left

💡 Hint

Common Mistakes

Using incorrect method names or variable names for retriever.

The method to get a retriever from the vectorstore is 'as_retriever'. This retriever is then passed to the RetrievalQA chain.

5fill in blank

hard

Fill all three blanks to run a query and print the answer from the RetrievalQA chain.

LangChain

query = '[1]'
result = qa_chain.[2](query)
print(result['[3]'])

Drag options to blanks, or click blank then click option'

AWhat is RAG?

Brun

Cresult

Danswer

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong method names or dictionary keys for the result.

The query is a question string. The method to run the chain is 'run'. The answer is accessed from the 'answer' key in the result dictionary.